[{"content":" Abstract # Starting from the simple objective of writing an optimized lexer we will grapple with their fundamental performance factor: branch prediction. We will investigate how lookup tables can improve (or hurt) the CPUs prediction capabilities with some unexpected results. Over time our design evolves by combining parsing and lexing into a single step in a way that is as fast as just a standalone lexer. A central aspect of the design will be a primitive parsing state machine from which the parser source code is generated.\nAfter the talk attendees will have a better understanding of some advanced branch prediction techniques and should have some new ideas for their present or future parsing endeavors.\n","date":"17 September 2026","externalUrl":null,"permalink":"/talks/cppcon2026/","section":"Talks","summary":"","title":"Writing High Performance Parsers Using State Machines","type":"talks"},{"content":" Abstract # We examine the canonical properties of general relativity on a light-front starting from the first order tetradic action. We focus on reducing the number of required variables through gauge fixing and solving constraints. This approach turns out to be quite fruitful as we are able to eliminate 4 out of the 6 internal rotation symmetries and a 3-dimensional subgroup of diffeomorphisms. We find that, contrary to the equal-time formulation, the resulting model naturally admits a non-vanishing Hamiltonian. Further our gauge choices allow us to give a closed form expression for the Dirac bracket in an appropriate basis. Last we review options to fix the remaining gauge symmetries to possibly obtain a model with only physical variables in the future.\n","date":"14 November 2023","externalUrl":null,"permalink":"/academic/master/","section":"Academic Work","summary":"","title":"Canonical Reduction of General Relativity on a Light-Front","type":"academic"},{"content":"Charge is an experimental memory-safe systems programming language, most similar to C++ and Rust. It is a personal research project in early development. The primary goals of the language are:\nProven memory safety \u0026hellip; : rigorous and semi-automatic proofs, enabled by formalizing reference guarantees, object invariants and more in the project\u0026rsquo;s own intermediate representation Chiral. \u0026hellip; that doesn\u0026rsquo;t get in the way: most code should not interact with the proof system directly. Writing manual proofs is the escape hatch from inherently safe borrow-checked semantics. Extensive proof system: advanced memory techniques and hacks that are possible in C++ should be possible in Charge too, when accompanied by a proof. Compact and powerful foundation: the language is built on few but general concepts, unifying aspects of both C++ and Rust. Related blog posts will appear here soon. For now checkout the Github repository for more details.\n","externalUrl":null,"permalink":"/projects/charge/","section":"Projects","summary":"An experimental memory-safe systems programming language","title":"Charge","type":"projects"}]